On Fri, Mar 20, 2009 at 12:46 PM, Neil Bartlett <[email protected]> wrote:
>
> One word of caution. It sounds like this approach would not work in a
> variety of classloading scenarios, such as J(2)EE or OSGi, since you
> make some assumptions about JAR files and the visibility of the
> application "classpath" etc.
Agreed -- I know well that Jython's method has some powerful limitations.

> I do a lot of work in OSGi and the proposal from John to get all
> classes in a package according to one classloader would probably work
> well, since OSGi strongly discourages "split packages", i.e. packages
> provided by more than one module and therefore classloader.
I think John's method could work well for us.  Given that the name is
"getLoadedClasses" I worry about classes that could be reached by the
class loader but have not yet been loaded.  I am on shaky ground here
I must admit, but my understanding is that classes are not guaranteed
to be loaded until they are used -- I presume that this means some
code somewhere has triggered a ClassLoader#loadClass -- so would
looking for them in this manner kick them into being loaded?

The javadoc for java.lang.ClassLoader#loadClass gives me hope:

Returns the class with the given name if this loader has been recorded
by the Java virtual machine as an initiating loader of a class with
that name.

-Frank

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "JVM 
Languages" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/jvm-languages?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to